Abstract 9458: Coronary CT Angiography for Re-Adjustment of Pre-Test Probability of Obstructive Coronary Artery Disease

Abstract

Backgrounds: Although various clinical algorithms have been proposed for defining pre-test probability of obstructive coronary artery disease (CAD), including Duke clinical score, these metrics suffer from selection bias of inclusion of predominantly high-risk population suitable for conventional coronary angiography. However, with noninvasive coronary CT angiography (CTA) comparative anatomy may also be available for a larger number of low and intermediate risk subjects, and may allow development of more inclusive algorithms. Methods: 3468 consecutive patients underwent CTA for suspected CAD (65.3±11.3 years, M:F: 1837:1811); patients with known CAD and those showing poor image quality were excluded from the study. Regression coefficients of age, gender, coronary risk factors including hypertension (HT), diabetes (DM), hyperlipidemia (HL) and smoking, history of cerebral infarction and chest symptoms (typical, atypical, or none) were weighed as predictors for CAD by multivariate analysis using logistic regression analysis against CTA findings. Risk score (K-risk) regarding the pre-test probability of CAD were calculated using each regression coefficients Results: Regression coefficients of age, gender, HT, DM, HL, smoking, history of cerebral infarction, chest symptom were 1.222, 0.039, 0.386, 0.768, 0.818, 0.385, 0.587, 0.431, respectively, and constant -6.404. The rate of concordance of K-risk and observation value was 82.8%. Area under curve of K-risk calculated by receiver operating characteristic curve was greater than that of Duke clinical score (0.769 vs 0.734 p=0.003). Conclusions: A K-risk score developed by comparison to CTA findings in the real-world population with suspected CAD allows for a superior assessment of pre-test probability of CAD. Further study for precision of it would be needed in a multicenter trial.